Given　poor　path　quality,　long　replanning　time,　and　slow　convergence　of　traditional　sampling　planning　algorithms　in　dynamic　environments,　an　optimal　dynamic　rapidly　exploring　random　tree*(D-RRT*)　algorithm　with　global　planning　combined　with　local　replanning　is　proposed.　First,　for　the　blind　search　problem　inherent　in　the　traditional　RRT　algorithm,　a　goal-biased　strategy　is　introduced　to　reduce　the　redundant　search　and　accelerate　the　convergence　of　the　algorithm.　Secondly,　a　triangular　inequality-based　inverse　optimization　strategy　is　introduced　to　optimize　the　path.　Then,　global　planning　combined　with　local　replanning　is　used　to　improve　the　real-time　performance　of　the　algorithm　for　dynamic　environments.　Finally,　D-RRT*　is　simulated　and　compared　with　the　traditional　sampling　algorithm　for　analysis,　and　the　experimental　results　verify　the　efficiency　and　stability　of　the　D-RRT*　algorithm　in　the　dynamic　environment.　©　2022,　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.